HOW BALLADS SPREAD 73

the identity of features in European ballad poetry, we shall proceed to
show that the earlier genre of ballads with refrain sprang from the same
primitive custom of dance, accompanied by improvised song, which
still exists in Greece and Russia, and even in the valleys of the Pyrenees.

The questions might all be answered in the affirmative (though
Perrault's Cinderella is perhaps not the happiest instance of time-
lessness), and not lead to the conclusion Lang announced. The
answers would refer to the isolated features of ballad poetry and
traditional literature, features which are to be found also in the
epical works he deemed aristocratic and personal; but the ballads
are not isolated features, but complexes of a definite type. It is at
this point we must remember that the ballad, for the duration of
this book, is a narrative song. Pure lyrics do not offer us a web of
circumstance such that we can recognize repetitions of the same
pattern without great difficulty. Their motifs are universal and
simple, and they are presented without elaboration. There is still
no reason to deny that each such song arose at a particular date
under the hand of a particular author; the fact that we do not know
either date or author does not imply that neither existed. Indeed,
the lyrics may be, though we do not know it, quite young; ever-
lasting reincarnations of the same simple notions. From the pure
lyric it is not possible to draw conclusions; but the narrative ballad
is quite different. We have already noticed that there exists a vast
mass of testimony as to the date of the genre in different European
countries, and the periods within which some particular pieces
have arisen. It is because these narratives are complexes of motifs
that this dating is possible; the motifs may be universal, but the
union established between them is temporal and distinctive. The
form also is generally defined with precision; one cannot mistake .
the form of a 'romance', Vise', 'junacka pesma*, or 'bylina1. This
form has its historical relation to other forms before and after, and
itself implies a chronology, though we may not know it. The
grouping of the motifs is too close to be explained by fortuitous
coincidence, and not infrequently gives some indication of the
circumstances attending on the original composition or the course
of migration.

Within any given area there is never a doubt but that two ballads
of identical content are forms of the same ballad, and not separate
spontaneous creations. The Three Ravens and the Twa Corbies are
forms of the same ballad tradition, though the one be English and

4615 T